The envelope glycoprotein of transmitted HIV is less glycosylated compared to viruses from later in infection. Glycans are added or shifted to escape strain-specific neutralizing antibodies, and in some cases these contribute to the formation of broadly neutralizing antibody epitopes. Thus, in some individuals, viral evolution may provide the stimulus for the development of broadly neutralizing antibodies. We hypothesize that early strain-specific neutralizing antibodies target exposed peptidic structures while later affinity matured antibodies from the same clonotype recognize the glycosylated epitope. A secondary hypothesis is that the inability of antibodies elicited in the RV144 vaccine trial to mediate neutralization is because they fail to recognize glycans. We will explore these hypotheses using well-characterized serial samples from HIV-infected women in the CAPRISA cohort and from the RV144 HIV vaccine trial. Serum samples will be run on glycan arrays to determine whether glycan-binding antibodies are preferentially found in infected women who later develop neutralization breadth and to assess overall levels of glycan reactivity in RV144 samples. The finding that V2 binding antibodies that recognize the K169 residue in the V2 region are an immune correlate in RV144, has led us to focus on infected women who develop broadly neutralizing antibodies that target this same residue and are glycan-dependent. We will use novel V1/V2 scaffolds that bind the glycan-reactive broadly neutralizing PG9 monoclonal antibody to study how V2 antibodies develop. Scaffolds will be modified in order to characterize antibody specificities and glycan-reactivity over time, and used in adsorption experiments to assess the neutralizing capacity of V2-directed plasma antibodies. We will define whether anti-V2 binding antibodies are qualitatively different in individuals who later develop broadly neutralizing anti-V2 responses, and whether the V2 specificities elicited in RV144 are similar to those elicited in natural infection. Finally, we will assess whether V2 binding antibodies are the precursors of broadly neutralizing V2 antibodies that develop in some HIV infected individuals by isolating V2 antigen-specific mAbs using single B cell sorting at early and late time points from selected CAPRISA women. These mAbs will be assessed for glycan-binding and functional activity including neutralization and inhibition of 47 binding. Antibody genes from these and other broadly neutralizing mAbs, recently isolated through B cell culture by our collaborators, will be analyzed to determine which features associate with the acquisition of neutralization breadth. Collectively, these data will reveal the pathway to neutralization breadth for anti-V2 antibodies, which may have important implications for understanding and improving on the RV144 results.